Displacement residual based DDM matching algorithm

• Published in: Science China. Information sciences Vol. 55; no. 9; pp. 2090 - 2101
• Main Authors: Zhang, Lin, Zhou, Zhong, Liu, Lin, Wu, Wei
• Format: Journal Article
• Language: English
• Published: Heidelberg SP Science China Press 01.09.2012; Springer Nature B.V
• Subjects: Algorithms; Architecture; Communications traffic; Computer Science; Computer simulation; DDM; Displacement; Distribution management; Information Systems and Communication Service; Matching; Networks; Payloads; Reduction; Research Paper; 位移; 匹配效率; 区域匹配算法; 数据分配; 残余; 网络负载; 高层体系结构; distributed simulation; region snapshot; high level architecture; data distribution management; region residual
• ISSN: 1674-733X; 1869-1919
• DOI: 10.1007/s11432-011-4427-0

High level architecture （HLA） is the prevailing standard for modeling and simulation. The data distribution management （DDM） service of HLA is defined for reducing the delivery of irrelevant data. The key in DDM implementation is the region overlap computation, i.e. the matching between update and subscription regions. Existing algorithms usnally make a compromise between region fidelity and network payload. This paper takes both the matching algorithm efficiency and bandwidth cost into account. The main contributions are： 1） illustrating the relationship between region changes and overlap changes, as helps reduce the number of region matching and then improves the total matching efficiency; 2） classifying region updates into two types of data expression, snapshot and residual. The network traffic will be reduced by transinitting only residual data instead of full region representations occasionally. Consequently, a region matching algorithm called displaeement residual-based DDM matching （DRBM） is proposed in the paper. Theoretical analysis, algorithm implementation and experiment evaluation are presented. Experinment results show that DRBM provides better matching performance and significant network payload reductions especially when there is a large number of changing regions.